UKPLab/sentence-transformers

Issues

• SentenceTransformerTrainer reset global step around step 127421 with no warning message

  #3145 opened 2 days ago by keeeeenw
  0

• grad_norm 0.0 while finetuning using group_by_label batch sampler

  #3130 opened 15 days ago by AmoghM
  2

• How to load ModernBERT model correctly?

  #3141 opened 2 days ago by Hannibal046
  3

• SigLIP-style loss for better DDP

  #3119 opened 21 days ago by kddubey
  7

• cannot import name 'is_torch_npu_available' from 'transformers'

  #3143 opened 4 days ago by LetsSolveAnalytics
  1

• cant load sentence-transformers/all-MiniLM-L6-v2 offline

  #3142 opened 5 days ago by wangguan1995
  2

• Differing behaviour with remote_code

  #3140 opened 7 days ago by Alicimo
  1

• How to set a proper batchsize when using CachedMultipleNegativesRankingLoss?

  #3134 opened 13 days ago by awmoe
  3

• How to avoid the long time waiting before start training?

  #3133 opened 3 days ago by awmoe
  3

• Memory leaked when the model and trainer were reinitialized

  #3136 opened 11 days ago by earthlovebpt
  3

• Batch sampler

  #3123 opened 17 days ago by nguyenvannghiem0312
  3

• MSMARCO training with SentenceTransformersTrainer instead of deprecated training scripts

  #3128 opened 16 days ago by sirCamp
  1

• multiple negative for MNRL

  #3139 opened 7 days ago by OnAnd0n
  5

• bug? cpu memory leak in model.encode when training in gpu

  #3138 opened 8 days ago by JazJaz426
  1

• import sentence_transformers出错

  #3135 opened 13 days ago by qsuzer
  1

• Inefficient loss calculation in cached losses

  #3107 opened 13 days ago by Marcel256
  2

• CrossEncoder gives OSError with newest Transformers version

  #3129 opened 15 days ago by Pringled
  5

• CrossEncoder .rank condition error in CrossEncoder.py

  #3124 opened 14 days ago by saeeddhqan
  6

• Changing Encoding Precision Yields error.

  #3121 opened 14 days ago by NovaBro
  7

• How to train a model with DDP for TSDAE

  #3132 opened 15 days ago by OnAnd0n
  1

• NanoBeirEvaluator takes in empty dataset

  #3103 opened 16 days ago by JINO-ROHIT
  4

• Core dumped in multi-cpu mode in container

  #3120 opened 19 days ago by jsirex
  1

• Cant train jina clip v2 due to CUDA out of memory error

  #3113 opened 23 days ago by httplups
  1

• Experience in Fine-Tuning a Cross-Encoder Model

  #3085 opened 20 days ago by nguyenvannghiem0312
  4

• missing docuementation for encode for image embedding models

  #3118 opened 21 days ago by KennethEnevoldsen
  1

• Loading private model with `trust_remote_code=True` fails

  #3115 opened 22 days ago by deklanw
  4

• adding new tokens to tokenizer without disturbing the base models embedding weight metrics of tokens

  #3116 opened 22 days ago by riyajatar37003
  3

• convert hugging face model to sentence transformer

  #3117 opened 22 days ago by riyajatar37003
  4

• save function does save modules.json incorectly for jina-embeddings-v3

  #3111 opened 23 days ago by guenthermi
  3

• Minor Difference in Embeddings Between Batch and Single Sentence Encoding

  #3109 opened 22 days ago by XYZliang
  2

• Breaking change for module loading in PyLate

  #3108 opened 23 days ago by NohTow
  0

• OSError: all-MiniLM-L6-v2 is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models

  #3106 opened 23 days ago by mai1x9
  3

• When using the SentenceTransformer.encode() method, the CPU usage reaches 100%

  #3091 opened 23 days ago by hello1534
  3

• Understanding how (hard) negatives in MNRL are used

  #3097 opened 24 days ago by HenningDinero
  4

• `CrossEncoder` is not pushed to cuda until predict is called, even if cuda is specified as device.

  #3078 opened 24 days ago by susnato
  5

• Error installing sentence-transformers with pip in Windows 10 pycharm professional-> Error: Dependencies conflicts most probobly

  #3077 opened a month ago by Umais-Adeed
  6

• NanoBeirEvaluator returns a mix of floats and numpy.float64 in the results.

  #3093 opened a month ago by JINO-ROHIT
  2

• Documentation Entry for `.encode` Method is Gone From Documentation Site

  #3094 opened a month ago by tedliosu
  1

• Parse arguments with HfArgumentParser

  #3090 opened a month ago by chocoded
  1

• SentenceTransformer._first_module() and issues created with bespoke architectures

  #3087 opened a month ago by slobstone
  0

• UnboundLocalError: local variable 'SentenceTransformer' referenced before assignment

  #3086 opened a month ago by riyajatar37003
  2

• BinaryClassificationEvaluator returns np.float32() and np.float64()

  #3075 opened a month ago by chr-werner
  2

• In Batch Negatives

  #3072 opened a month ago by riyajatar37003
  24

• Can't start training with a SentenceLabelDataset due to error `'SentenceLabelDataset' object has no attribute 'column_names'`

  #3082 opened a month ago by HenningDinero
  2

• Best way to log specific metrics e.g cross-entropy/accuracy during training

  #3080 opened a month ago by HenningDinero
  5

• query adapter native in training

  #3084 opened a month ago by achibb
  3

• Same order of training samples with `NoDuplicatesBatchSampler`

  #3069 opened a month ago by antigregory
  4

• Weights of models trained with TSDAE

  #3067 opened a month ago by xboldt
  2

• Deduplication integrated in CachedGISTEmbedLoss

  #3063 opened a month ago by yjoonjang
  2

• Add `normalize_embeddings` Argument to `SentenceTransformer` for Simplified Embedding Normalization

  #3064 opened a month ago by AIMacGyver
  0